How Data Visualization Allowed an Ed-Tech Startup to Boost Conversions: Refocus Case

• The Client aimed to foster data-based decision-making to enhance their sales and product development. However, they lacked internal resources to build a data analytics infrastructure, come up with the right tools, and set up the processes.
• Refocus looked for a reliable partner with extensive hands-on expertise and a focus on details and business goals. Since Valiotti met all the requirements, we were hired to perform the task.

While starting to work on the project, we had to struggle through a technical challenge: Refocus’s team modified data sources and didn’t notify us, which caused a lot of trouble for our data analysts. How did we rule this out? We had to ask Refocus not to alter anything that may be streamed into the data pipeline.

1. Scalable data analytics infrastructure

After analyzing the Client’s business goals and data collection requirements (i.e., data sources, the required attributes), we came up with a tech stack that included:

• Python scripts
• The Hetzner server for storing the data collection scripts
• Airflow for workflow management
• Google BigQuery as a data warehouse
• Tableau for data visualization

With the help of Airflow and Python scripts, the data is exported, processed, and stored in the database via API. Hence, all the data collected from tables is in Google BigQuery. Airflow, which is based on a virtual machine, manages the workflows.

All the data is then visualized in the Tableau reports.

2. Numerous regularly updated dashboards for actionable insights

The Client can build both generic and specialized reports to make data-driven decisions.

For example, here is an SLA report.

3. Optimized lead allocation among employees for improved conversions

We analyzed various lead funnels and reallocated them to increase the conversion rate from X% to Y%. This allowed the Client to speed up the conversion from a new deal to a sale.

4. Detecting the top reasons for refunds

The dashboard for students’ refund requests determines the most common reasons for refunds. Now, the Client can work them through to minimize the amount, thus profiting from the data available to them.

5. Spotting the most complicated tasks

By analyzing dashboards that illustrate students’ performance, the Client managed to monitor the progress of each student as well as discover the most difficult tasks and lessons. The insights allow for data-based product development, which may increase course completion rates in the future.